WT-1 positivity has previously been noted in nonneoplastic endometrial stroma. In this study we examined WT-1 expression in endometrial stromal
neoplasms to ascertain whether these
tumors are immunoreactive and whether this antibody might be of value in the diagnosis of these lesions. We also stained cases of cellular and highly cellular
leiomyomas to investigate whether WT-1 might be of value in distinguishing these from an endometrial stromal
neoplasm. We compared WT-1 staining with CD10,
desmin, alpha smooth muscle actin, h-
caldesmon, and AE1/3, many of these
antibodies being commonly used to distinguish between an endometrial stromal and a smooth muscle phenotype. Cases of ESN (n = 5), low grade ESS (n = 14), and cellular or highly cellular
leiomyoma (n = 14) were stained with the aforementioned
antibodies. Cases were scored on a scale of 0 to 4+, with 4+ cases exhibiting positivity of >50% of cells. Sixteen of 19 endometrial stromal
neoplasms were positive with WT-1, most (14 of 16) with 4+ positivity. Staining was nuclear (5 cases), cytoplasmic (5 cases), or combined nuclear and cytoplasmic (6 cases). All endometrial stromal
neoplasms exhibited 4+ staining with CD10. Staining for alpha smooth muscle actin was present in most cases (14 of 19) and
desmin and h-
caldesmon were positive in a smaller number of cases (8 and 2 respectively). There was 4+ positivity with
desmin in only 1 case. The 2 cases that were h-
caldesmon positive both exhibited 1+ staining (<5% cells positive). Six cases were positive with AE1/3, 1 with 4+ staining. Leiomyomatous
neoplasms always exhibited 4+ staining with
desmin and alpha smooth muscle actin and in most cases (12 of 14) with h-
caldesmon. The other 2 cases exhibited 2+ positivity. Most cases (12 of 14) were positive with WT-1 (7 of 14 with 4+ staining) and CD10 (5 of 14 with 4+ positivity). One case was positive with AE1/3. We conclude that diffuse WT-1 positivity is characteristic of endometrial stromal
neoplasms and that this may be of value in diagnosis. However, WT-1 is of limited use in the distinction between an endometrial stromal and a cellular leiomyomatous
neoplasm because many of the latter are also positive. This study confirms the value of h-
caldesmon in the distinction between an endometrial stromal
neoplasm (almost always h-
caldesmon negative) and a cellular leiomyomatous
neoplasm (h-
caldesmon positive). Although CD10 is positive in endometrial stromal
neoplasms, the commonly observed immunoreactivity of cellular and highly cellular
leiomyomas with this antibody limits its diagnostic usefulness.
Desmin is useful as all leiomyomatous
neoplasms exhibited diffuse positivity, whereas only a small number of endometrial stromal
neoplasms were focally positive and only 1 case exhibited 4+ positivity. Smooth muscle actin is of limited value since most
neoplasms studied were positive. The overlapping immunophenotype of endometrial stromal and leiomyomatous
neoplasms may reflect the origin of both cell types from a common progenitor within the uterus.